Comparative Performance of Holt-Winters and Prophet in Forecasting Electricity Demand in Distribution Systems and Public Lighting
DOI:
https://doi.org/10.61236/qd0p8f87Keywords:
electricity demand, distribution networks, public lighting, forecasting, Holt-Winters, ProphetAbstract
Accurate electricity demand forecasting supports operational planning and energy-efficiency decisions in distribution systems. This study compares the additive Holt-Winters and Prophet models applied to public lighting demand in Cotopaxi, Ecuador, using an institutional inventory of 52,365 luminaires from 2012 to 2020. The inventory was transformed into a monthly theoretical energy series through data cleaning, calendar-based energy calculation and a dynamic replacement rule for the transition from mercury and sodium lamps to LED technology. The models were implemented in Python and evaluated with MAPE, RMSE and R². Holt-Winters obtained lower short-term error, whereas Prophet better represented the 2019 LED structural change; therefore, both models are useful for complementary operational and strategic planning.
Downloads
References
[1] C. Á. Fierro Torres, V. H. Castillo Pérez y C. I. Torres Saucedo, “Análisis comparativo de modelos tradicionales y modernos para pronóstico de la demanda: estado del arte,” Revista Técnica Energía, 2024.
[2] E. J. Mejía Vásquez y S. Gonzales Chávez, “Predicción del consumo de energía eléctrica residencial de la Región Cajamarca mediante modelos Holt-Winters,” Ingeniería Energética, 2021.
[3] M. Muñoz García, D. Peña Banegas, J. Murillo Oviedo y F. Ortega Loza, “Análisis de los factores que condicionan la demanda energética en la Región Costa del Ecuador,” Revista Científica, 2022.
[4] I. O. Anaguano Pérez y F. F. Sandoya Sánchez, “Métodos basados en los modelos ARIMA estacional y Holt-Winters para la previsión de la demanda eléctrica a ultra corto plazo,” Revista Politécnica, 2020.
[5] J. W. Taylor, “Triple seasonal methods for short-term electricity demand forecasting,” European Journal of Operational Research, vol. 204, no. 1, pp. 139-152, 2010.
[6] E. S. Gardner Jr., “Exponential smoothing: The state of the art,” Journal of Forecasting, vol. 4, pp. 1-28, 1985.
[7] S. J. Taylor and B. Letham, “Forecasting at scale,” The American Statistician, vol. 72, no. 1, pp. 37-45, 2018, doi: 10.1080/00031305.2017.1380080.
[8] R. J. Hyndman y G. Athanasopoulos, Forecasting: Principles and Practice, 2nd ed. Melbourne, Australia: OTexts, 2018.
[9] D. C. Montgomery, C. L. Jennings y M. Kulahci, Introduction to Time Series Analysis and Forecasting, 2nd ed. New Jersey, NJ, USA: John Wiley & Sons, 2015.
[10] G. E. P. Box y G. M. Jenkins, Time Series Analysis: Forecasting and Control. San Francisco, CA, USA: Holden-Day, 1976.
[11] O. Trull Domínguez, “Predicción a corto plazo de la demanda horaria de energía eléctrica en España mediante modelos optimizados de Holt-Winters múltiple-estacional,” Trabajo de Fin de Máster, Univ. Politécnica de Valencia, Valencia, España, 2024.
[12] J. M. Sánchez López, “Pronóstico de consumo eléctrico y detección de anomalías mediante machine learning,” Trabajo de Fin de Máster, Escuela Técnica Superior de Ingeniería, España, 2024.
[13] E. Romero Depablos, “Predicción del volumen neto de los desvíos generación/demanda (MW) de la red eléctrica de España,” Trabajo de Fin de Máster, Univ. Europea de Madrid, Madrid, España, 2025.
[14] A. Toapanta Lema, “Comparación de modelos de regresión para la eficiencia en el consumo de electricidad en las escuelas ecuatorianas: un caso de estudio,” Repositorio UISEK, Ecuador, 2020.
[15] C. C. Holt, “Forecasting seasonals and trends by exponentially weighted averages,” O.N.R. Memorandum, vol. 52, Carnegie Institute of Technology, 1957.
[16] P. R. Winters, “Forecasting sales by exponentially weighted moving averages,” Management Science, vol. 6, no. 3, pp. 324-342, 1960.
[17] F. X. Diebold and R. S. Mariano, “Comparing predictive accuracy,” Journal of Business & Economic Statistics, vol. 13, no. 3, pp. 253-263, 1995, doi: 10.1080/07350015.1995.10524599.
[18] Secretaría de Energía, “Acuerdo por el que se emite el Manual de Pronósticos,” Diario Oficial de la Federación, México, nov. 23, 2017.
[19] Agencia de Regulación y Control de Electricidad, “Regulación No. ARCONEL-006/18: Prestación del Servicio de Alumbrado Público General,” Ecuador, 2018.
[20] Statsmodels Developers, “statsmodels.tsa.holtwinters.ExponentialSmoothing,” documentación oficial, 2026. [En línea]. Disponible en: https://www.statsmodels.org/. Consultado: 2 jul. 2026.
[21] Meta Open Source, “Prophet: Forecasting at scale,” documentación oficial, 2026. [En línea]. Disponible en: https://facebook.github.io/prophet/. Consultado: 2 jul. 2026.
[22] Gaón-Escobar K, Quinatoa-Caiza C, Proaño-Maldonado X, Mullo-Pallo M, "Advanced Statcom Integration Through Hybrid Co-Simulation for Stability Improvement in Power Systems," WSEAS Transactions on Power Systems, vol. 20, pp. 493-502, 2025, DOI:10.37394/232016.2025.20.39
[23]“Estimación Paramétrica del Modelo de Carga de Recuperación Exponencial Utilizando Mediciones Sincrofasoriales”, re, vol. 22, no. 2, pp. PP. 65–74, Jan. 2026, doi: 10.37116/revistaenergia.v22.n2.2026.735.
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Ciencias de la Ingeniería y Aplicadas
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
The authors who publish in this journal agree to the following terms:
- Creative Commons Attribution-NonCommercial License allows others to share the work with acknowledgment of authorship of the work and initial publication in this journal.
- Authors may separately establish additional agreements for the non-exclusive distribution of the version of the work published in the journal (for example, placing it in an institutional repository or publishing it in a book), with an acknowledgment of its initial publication in this journal.
- Authors are permitted and encouraged to disseminate their work electronically (for example, in institutional repositories or on their own website) before and during the submission process, as it can lead to productive exchanges, as well as further citation. earliest and largest of published works (See The Effect of Open Access) (in English).
